accuracy of thematic maps / implications of choropleth symbolization
Mapping census data - approaches and issues · Bivariate Choropleth Maps • Most Choropleth maps...
Transcript of Mapping census data - approaches and issues · Bivariate Choropleth Maps • Most Choropleth maps...
Mapping census data – approaches
and issues
James Crone
Edinburgh University / UK Data Service
An Introduction to Geographical Data
Visualisation
JISC Manchester
Thursday 16th May 2019
Census data available through UKDS
(Small Area) Census aggregate data
Geospatial Data / GIS
• GIS is a system designed to capture, store, manipulate, analyse, manage and present geospatial data.
• Geospatial data models some aspect of the real world whether that be the natural, built or socio-economic world
• Raster data models the world as a continuous grid of equally sized cells
• Vector data models the world as point; line or polygon features.
• Mostly when we are dealing with census data we are dealing with vector data although raster data can be used e.g. for population surfaces.
Visualising census data
Mapping census data as a Choropleth map
• A thematic map in which areas are shaded or patterned in proportion to the measurement of the statistical variable being displayed on the map.
• Choropleth maps provide an easy way to visualize how a measurement varies across a geographic area or to show the level of variability within a region.
% of people working > 49hrs per week as recorded in the 2011 census
Pick census variable and output Geography
• Small area aggregate census data is available, output at different levels of geography:
Country
Local Authority
MSOA
LSOA
OA
• Due to disclosure control, some small area aggregate census data is only available at certain levels of geography. Data at Local Authority level may not be available at OA level.
• Where Data is available at different levels of geography, different patterns might be shown depending on the level of geography selected.
Prepare data for mapping
• When constructing a choropleth map from polygonal census boundaries
which are usually of different sizes, displaying raw counts of census
variables should be avoided.
• Instead the census variables being displayed should be normalized.
• Option 1 – Normalize the census variable being mapped by dividing it by
the total geographic area. This expresses the variable as a density.
• Option 2 – Normalize the census variable being mapped by dividing it by
the total population size (people or households) in that area.
Statistical Classification• Classification takes a large number of
observations and groups them into a smaller number of data ranges or classes.
• This makes it easier to spot patterns and understand the data compared with looking at all the variance of the data at once.
• Different classification methods are available. Some of these include:
• Equal Interval
• Quantile
• Natural Breaks
• Manual
• No classification method is right or wrong. Choice of classification method should be based on the characteristics of the data.
Style the map by applying a colour ramp to classes
Diverging
-30 -20 -10 0 10 20 30
Qualitative
A B C D E F G
Sequential
10 20 30 40 50 60 70
Bivariate Choropleth Maps
• Most Choropleth maps only
display a single variable
• A Bivariate Choropleth map
combines data from 2
variables at the same time
• Here we can see in a single
map BOTH where the vote
and access to healthcare
was high or low or a mix of
the 2 variables.
Two limitations of Choropleth Maps
• Choropleth maps imply
that the population is
distributed uniformly
across the extent of the
polygon (census zone).
• Small polygons are often
hidden by larger areas.
• Alternatives to Choropleth
Maps?
Dasymetric Maps
• Modify existing Choropleth map using additional geospatial data such as residential areas or buildings
• DataShine is a form of Dasymetric map in that area based census data has been redistributed to building features within each census area.
• This helps with the problem of the choropleth implying that population is uniformly distributed across polygons
Cartograms
• In a Cartogram, the polygon geometry is distorted or reshaped according to the variable being mapped (rather than being shaped according to the land area of the geography being shown)
• This helps with the problem of small areas being hidden by larger areas.
• Cartograms are a form of map projection
• Types of Cartogram:
Non-contiguous Cartograms
Contiguous Cartograms
Dorling Cartograms
Use of Cartograms
QGIS Desktop GIS Application
https://www.qgis.org
Use QGIS to map data Exercises• Use QGIS to:
• Exercise 1: Create a Choropleth map
• Exercise 2: Create a Bivariate Choropleth map
• Exercise 3: Create a Contiguous Cartogram
• Exercise 2 and 3 require additional data to be downloaded, go to:
https://bit.ly/2VE9bem
and download the UKDSGeoDataViz.zip file to your working folder.
• Unzip the contents of the UKDSGeoDataViz.zip file
• Start with Exercise 1. If / when you finish Exercise 1, take a look at Exercise 2 or 3 (or both!)
• If you get stuck / have questions put your hand up